GAN-based Chest X-ray Image Synthesis

Project Overview

This project implements a Generative Adversarial Network (GAN) to generate synthetic chest X-ray images for data augmentation and research purposes. The goal is to tackle data scarcity in medical imaging by creating realistic X-ray images of both normal and pneumonia cases.

Features

• Generates synthetic chest X-ray images to improve dataset balance.
• GAN architecture with a generator and discriminator model.
• Ability to augment training data for medical AI models.
• Useful in addressing challenges such as data imbalance and data limitations in healthcare.

Dataset

The Chest X-ray dataset used for this project can be found on Kaggle.

Getting Started

Prerequisites

• Python 3.x
• TensorFlow 2.x
• Keras
• NumPy
• Matplotlib

Install the required packages using the following command:

pip install -r requirements.txt